The gap between technological innovation and regulation is becoming one of the most significant workplace compliance challenges facing businesses today. Facing labor shortages, rising costs, and growing competitive pressure, employers are increasingly turning to artificial intelligence for efficiency, scalability, automation, and cost reduction. Yet those benefits come with substantial legal and reputational risks.
Businesses are discovering that AI tools can create exposure in unexpected ways. Automated hiring systems may unintentionally screen out older applicants. Productivity-monitoring tools can disproportionately impact employees with disabilities. Generative AI platforms may expose trade secrets or jeopardize attorney-client privilege. The question is no longer whether AI can improve workplace efficiency, but whether it can be deployed responsibly and defensibly.
The Growing Patchwork of State Laws
Many lawmakers view AI regulation similarly to earlier workplace and consumer protection movements in which states acted before Congress. As a result, employers now face a fragmented compliance environment that varies significantly by jurisdiction.
One prominent example is New York City’s Local Law 144, which requires businesses that use automated employment decision tools to conduct bias audits and to provide notice to applicants and employees. Similar laws have emerged in Illinois, Colorado, California, and New Jersey. Although these laws differ, they generally focus on algorithmic transparency, bias testing, notice requirements, human oversight, documentation obligations, and anti-discrimination enforcement.
For multi-state employers, compliance can be particularly difficult. A single AI hiring platform may simultaneously trigger multiple, and sometimes conflicting, state requirements.
The growing intersection of AI and employment discrimination law is illustrated by Mobley v. Workday, Inc., 740 F. Supp. 3d 796 (N.D. Cal. 2024). There, the plaintiff alleged that Workday’s AI-driven applicant screening system disproportionately excluded applicants based on race, age, and disability. Workday argued it could not be liable because it functioned solely as a software vendor, not as an employer or employment agency. The court rejected that argument at the pleading stage, holding that Workday could potentially qualify as an “agent” of an employer when hiring authority is delegated to algorithmic systems.
The case also reflects growing judicial skepticism toward “black box” AI systems with decision-making processes that cannot be adequately explained or audited.
The Federalization of AI Policy
As states move aggressively to regulate workplace AI, the federal government has signaled a different approach. On January 23, 2025, the Trump administration issued Executive Order 14179, Removing Barriers to American Leadership in Artificial Intelligence. The order emphasizes federal control over AI policy and criticizes state AI laws as burdensome to innovation.
The Executive Order establishes a federal AI Litigation Task Force to review and potentially challenge state AI laws believed to conflict with federal objectives. For businesses, this shift presents both opportunities and risks. A more centralized framework could reduce compliance fragmentation for employers operating nationwide. At the same time, broader federal preemption could weaken state-level civil rights protections just as AI systems increasingly influence hiring, compensation, discipline, and other critical workplace decisions.
The Executive Order also raises concerns about the future of over 50 years of disparate impact jurisprudence. First formally recognized by the Supreme Court in Griggs v. Duke Power Co., 401 U.S. 424 (1971), disparate impact claims allow plaintiffs to challenge facially neutral policies that disproportionately harm protected groups, even absent intentional discrimination. The Griggs decision fundamentally reshaped employment discrimination law by recognizing that discrimination can result not only from overt bias but also from structural systems that perpetuate inequality.
The disparate impact doctrine is especially important in the AI context because algorithmic systems are often trained on historical datasets that may reflect longstanding inequities. Eliminating disparate impact as an analysis and audit tool in the AI context could entrench biased datasets and potentially legitimize this bias as a truthful and accurate reflection of worker attributes, qualifications, and other employment factors.
What Businesses Can Do Now
Employers implementing AI tools should consider several immediate best practices:
- Conduct meaningful vendor due diligence;
- Negotiate strong contractual protections regarding confidentiality, indemnification, and ownership of training data;
- Protect privileged and confidential information from disclosure through third-party AI systems;
- Maintain meaningful human oversight over significant employment decisions; and
- Create cross-functional AI governance committees involving legal, HR, compliance, IT, and executive leadership.
Businesses should also recognize that AI governance increasingly affects reputation, employee morale, investor confidence, and public trust.
Conclusion
Technological disruption has always tested the law’s ability to adapt. AI may transform how companies hire, manage, and evaluate employees, but it does not alter the core principles underlying responsible employment practices. Fairness, accountability, and transparency remain essential. Employers that build their AI strategies around those principles will be best positioned to manage risk, maintain trust, and thrive in the evolving workplace.
This article may be reproduced, in whole or in part, with the prior permission of Thompson Hine LLP and acknowledgement of its source and copyright. This publication is intended to inform clients about legal matters of current interest. It is not intended as legal advice. Readers should not act upon the information contained in it without professional counsel.
This document may be considered attorney advertising in some jurisdictions.
© 2026 THOMPSON HINE LLP. ALL RIGHTS RESERVED.